spark中基于列的hive合并数据

8cdiaqws  于 2021-06-28  发布在  Hive
关注(0)|答案(1)|浏览(520)

我在配置单元表中有以下格式的数据。

user |  purchase | time_of_purchase

我想把数据放进去

user | list of purchases ordered by time

如何在pyspark或hiveql中执行此操作?
我曾尝试在配置单元中使用collect\u list,但它无法按时间戳正确保留顺序。
编辑:按kartikkannapur的要求添加样本数据。这是一个示例数据

94438fef-c503-4326-9562-230e78796f16 | Bread | Jul 7 20:48
94438fef-c503-4326-9562-230e78796f16 | Shaving Cream | July 10 14:20
a0dcbb3b-d1dd-43aa-91d7-e92f48cee0ad | Milk | July 7 3:48
a0dcbb3b-d1dd-43aa-91d7-e92f48cee0ad | Bread | July 7 3:49
a0dcbb3b-d1dd-43aa-91d7-e92f48cee0ad | Lotion | July 7 15:30

我想要的输出是

94438fef-c503-4326-9562-230e78796f16 | Bread , Shaving Cream
a0dcbb3b-d1dd-43aa-91d7-e92f48cee0ad | Milk , Bread , Lotion
34gzjxbg

34gzjxbg1#

一种方法是
首先创建一个配置单元上下文并将表读取到rdd。

from pyspark import HiveContext
purchaseList = HiveContext(sc).sql('from purchaseList select *')

然后处理rdd

from datetime import datetime as dt
purchaseList = purchaseList.map(lambda x:(x[0],[x[1],dt.strptime(x[2],"%b %d %H:%M")]))
purchaseByUser = purchaseList.groupByKey()
purchaseByUser = purchaseByUser.map(lambda x:(x[0],[y[0] for y in sorted(x[1], key=lambda z:z[1])]))
print(purchaseByUser.take(2))

输出

[('94438fef-c503-4326-9562-230e78796f16', ['Bread', 'Shaving Cream']), ('a0dcbb3b-d1dd-43aa-91d7-e92f48cee0ad', ['Milk', 'Bread', 'Lotion'])]

将rdd另存为新配置单元表

schema_rdd = HiveContext(sc).inferSchema(purchaseByUser)
schema_rdd.saveAsTable('purchaseByUser')

有关读取和写入配置单元表的信息,请参阅stackoverflow问题和spark文档

相关问题